R&D Goes Global

Back in the 1980s, politicians and pundits mourned the migration of blue-collar jobs from advanced industrial countries to emerging economies. More recently, they fretted over the flow of low-level white-collar jobs. In both cases, their one piece of advice for the advanced countries was to do what they did best: innovation.

Think again. Now even innovation appears to be fleeing to low-cost countries, as they seemingly leapfrog the traditional stages of development. Clearly, we are at the beginning of a fundamental shift in where R&D is performed.

Yet you won’t see a sign of that shift in this year’s R&D 100 list, compiled for IEEE Spectrum by Standard & Poor’s. High-cost countries continue to dominate the list, with the United States, Japan, Germany, and the United Kingdom accounting for more than two-thirds of the firms and more than three-quarters of the R&D spending. Not a single low-cost country has a firm in the top 100, including those based in China or India, even though those two countries are far and away the biggest new magnets for R&D investment.

The reason is simple: the spending trend flies below our radar, which scans the reports of publicly traded multinational corporations. Though ever more of them—including all top 10 firms on our list—are offloading R&D work to their subsidiaries in India and China and still others are subcontracting the work to local vendors, the spending still shows up on the parent companies’ books.

Certainly, there is no regulation that forces firms to break out the locations of R&D spending in their reporting. Just as certainly, there are competitive and even political motives to keep the locations hush-hush. Nobody wants to hear rabble-rousing engineers in the richer countries grousing about the ”giant sucking sound” made by Asia as it slurps up some of the profession’s highest-paid jobs.

In the meantime, you can still learn a lot from the R&D 100, because it tells you whom the innovators are working for, if not where they are working. And, as usual, the highest spending firms are in the automotive and pharmaceutical sectors, each with four firms. Rounding out the top 10 are software giant Microsoft and telecom powerhouse Nokia.

General Motors increased its R&D spending by a whopping 22.7 percent, in spite of the troubled firm’s nearly 13 percent drop in sales. It thus vaulted to second place, up from ninth last year. Nokia cracked the top 10 for the first time, rising to fifth place from 14th; this increase largely reflects the R&D component of its joint venture with Siemens, called Nokia Siemens Networks. Volkswagen has continued its steady ascent moving to 8th fromto last year’s 4th place and from 25th place back in 2002.

The changes in rankings also reflect currency fluctuations, especially the significant weakening of the value of the U.S. dollar over the past few years. Just four years ago, the U.S. dollar was on par with the euro; now it’s worth only about three-quarters of a Euro. And since our table is expressed in U.S. dollars, companies based in countries where the currency has strengthened, like those using the euro, will appear to be spending more than is actually the case. For instance, if the U.S. dollar were on par with the euro, as it was just five years ago, Nokia would be ranked behind IBM (No. 15 on the list as it now stands), instead of 5th. Whether any particular increase in spending reflects additional research activities or simply exchange rate fluctuations is impossible to say, at least without more detailed information about where the company in question does its R&D.

Microsoft moved up two slots, to ninth place, but even so it’s spending nearly a tenth less than the US $7.8 billion it spent in 2003, when it topped the leaderboard. In that period the company’s sales grew by 41 percent, which means that the ratio of R&D spending to sales—called R&D intensity—declined to 13.9 percent from 21.1 percent.

Automaker Daimler dropped to 22nd from second place largely on the divestiture of Chrysler, which doesn’t appear on the list because it’s privately held. Siemens dropped to 19th place from fourth because of a 32.3 percent cut in R&D this cut, however, appears to be just an artifact of the way it accounted for its joint venture with Nokia. In fact, Siemens reports that it increased its R&D spending by 300 million euros between 2006 and 2007.

Fast-growing Google continued its ascent up the list, to number 59, increasing its spending by 73.7 percent, significantly more than the 56.5 percent growth in its sales.

It would be great if all the big spenders broke their spending down by country, but perhaps they shouldn’t be faulted for the lapse—the magnitude of R&D outsourcing has only recently been noticed. The first really striking news came in 2006, when the Paris-based Organisation for Economic Co-operation and Development announced that the Chinese government’s R&D spending was second only to that of the U.S. government.

The same trend was documented in the private sector when Jerry and Marie Thursby, professors of management at the Georgia Institute of Technology, surveyed 248 R&D managers of U.S. and European multinational corporations. The managers said they’d recently opened or were planning to build more key R&D facilities in China than in the U.S. Managers expected R&D employment to rise in India and China, and many of them expected it to decline in the United States. A similar survey of 300 executives, conducted last year by The Economist , found that more chose India as the best overall location for R&D outside their home countries, followed by the United States and China. Canada was a distant fourth.

There are a number of reasons why this shift of R&D work is happening now. First off, multinationals need to do research in the countries where they sell to adapt products to local markets, a practice known as localization. Indeed, segments of these local markets are approaching a scale that until now had been seen only in developed countries. Also, to gain market access they need to play along with the policies that local governments have instituted to modernize their countries, partly for the prestige of it and partly for the independence from foreign know-how that it promises to deliver. And the low-cost countries need to move up from manufacturing to services to brainwork for the same reason Western countries did years ago—to ensure against the advent of even lower-cost competitors elsewhere. Finally, many of these investments were attracted by the lower costs of labor and capital in the emerging economies. Workers in Google’s Bangalore R&D center earn $30 000, a royal sum there but close to poverty levels for Silicon Valley. Government subsidies and tax policies often reduce the physical plant and land costs.

However, there is another, more obvious reason for the shift: China and India have been building capabilities. China, in particular, is building its R&D infrastructure at a breakneck pace. The country’s R&D spending has increased by an average annual rate of nearly 19 percent since 1995, according to the OECD Reviews of Innovation Policy: China , published in August.

Denis Fred Simon, a professor of international affairs at Penn State University, who specializes in analyzing China, says it has ”emerging pockets of excellence.” He says that China’s government reports that foreign multinationals have established 1160 R&D centers, up from just 30 in 1999. Some of these centers are doing cutting-edge work. And indigenous innovation, a priority in China’s most recent five-year plan, appears to be taking hold. Solar energy is one example, with companies like Suntech; pharmaceuticals are another, with Wuxi AppTec; telecommunications is a third, with Huawei. Among the foreign multinationals driving such work, Simon singles out General Electric, whose Shanghai facility is doing world-class development of algorithms for three-dimensional medical imaging, and Intel, which is building a $2.5 billion chip fab in Dalian, China.

In October, IBM announced that it was opening an R&D center in Shanghai, its second in China, and that it would eventually hire some 100 researchers to work on applications for the Internet and small businesses. IBM hadn’t opened such a center anywhere in the world since 1995.

Some analysts argue that China’s innovation capacity is merely a facade, and that India has the greater potential. China’s plethora of patents, papers, and technology parks are a mere shell without the rule of law, contends Vivek Wadhwa, an entrepreneur with adjunct appointments at Duke and Harvard universities. He says that in the long run India, with its tradition of independent courts, will reassure foreign investors that the intellectual property they bring to the country will be improved—and not pirated. That, he says, is one reason that Motorola was prepared to do some 40 percent of the development of its Razr phone in India, and why Boeing has outsourced so much R&D on airliner bodies and interiors to India’s tech services powerhouse, HCL Technologies. Wadhwa, like many other observers, says that R&D operations in India are making products for the global market, while R&D operations in China are mostly for localization efforts and some incremental innovation.

Even HCL is chary of China’s pirates. ”We have a center ourselves in China, but before we move any work to China we need to get our customers’ approval, because of their sensitivity to IP security,” says Sandeep Kishore, a senior vice president who heads HCL’s North American manufacturing.

HCL sold $512 million in R&D services in the year ending 30 June 2008, about a quarter of its total revenues, and practically every cent of the half-billion was for work done for overseas customers, appearing on their books as their own R&D costs. Half a billion wouldn’t place it on the R&D 100, now a billion-dollar club, but still, it’s 40 percent more than HCL did the previous year—which was 40 percent more than the year before that. If HCL can keep up this pace of growth, in six more years, its R&D budget will top Microsoft’s.

Of course, such a straight-line extrapolation isn’t realistic. At some point, the rapid growth rates of innovation work moving eastward will slow. Denis Simon says that both India and China are finding that it’s getting harder to recruit the people they need. In the early days—like, oh, six or seven years ago—they could woo their own expatriate engineers back from Silicon Valley or wherever else they’d landed. But there are only so many such returnees, as they’re called, which means future expansion will increasingly depend on locally trained talent.

China is indeed graduating huge numbers of engineers—two to three times as many as the United States produces. But according to Simon, just ”20 percent of that is really usable; a lot of the others are technicians. They haven’t the depth, and they won’t be innovators.” That’s why China is scouring the world for expatriates and working hard to entice them back.

”India was like that three to five years ago but no longer,” says Wadhwa. ”In fact, if [the expats] have spent more than seven years abroad, the companies now think of them as not fit for India anymore. Training here is better than in the West.”

Much of that training is in-house. HCL, for example, runs a finishing school for all its hires. The course lasts anywhere from 12 to 18 months, depending on the employee’s level of preparation, and all the instruction centers on what the company needs rather than what a university engineering department might prefer to teach.

To take their places among the true big boys of R&D, Chinese and Indian firms will have to do more than just add people, pour money into R&D, and attract foreign investment. Creating the proper ”ecosystem” for indigenous innovation requires many different elements, and they all have long gestation periods so it’s difficult to evaluate the return on investments. It will take us a long time to figure out the impact of all those spanking new R&D centers that have begun to dot Asia.

It may be quite a few years, then, before firms from India, China, Brazil and other low-cost countries begin to appear among the R&D 100. But even now, that list can be understood only in the light of a migration of innovation work to those countries.